Verba

Inspiration

In the past, I've dozed off during class and missed out on the notes. I wanted to come up with a solution for the times when I haven't had enough sleep to still get the notes.

What it does

It takes input from the users microphone and transcribes the information with timestamps in real-time.

How we built it

I used Next.js, TypeScript, Huggingface, and Supabase

Challenges we ran into

I ran into model optimization to where it took over a minute to transcribe what I said.

Accomplishments that we're proud of

I'm proud of optimizing the model to where it doesn't take as long as before.

What we learned

I learned about model optimization and creating a better user experience through faster loading times.

What's next for Verba

To go beyond the MVP and create user suggested features!

Built With

• huggingface
• next.js
• supabase
• typescript